Geo,

Depending on the output resistance of the BJT at high voltages (a number found in the datasheet or? with a curve tracer) performance might suffer w/o the cascode. This is because, in addition to enabling functionality of a LV BJT in this circuit, the cascode does another thing: it boosts the output resistance driving the 1M load resistor, thus boosting the open loop voltage gain of the gainstage (think "opamp") before the resistor feedback loop is closed.

Thus, the cascode helps boost the loop gain which is instrumental in determining the accuracy of regulation of the output voltage.

So, I am saying that a 1000V BJT will give functionality, but perhaps not as good regulation at the output.....

However, it might be "good enough." I don't know that yet. It might also be more economical.......depending on the price of 1000V small-signal BJTs..... I don't know.......

rogerw

?

"

(If you had a 1000V BJT you would not need the cascodes for mere functionality.)

rogerw"

?

Yep, that's been the dilemma?here, trying to find a way to transfer the stable "Variable Zener"? LM317? correction: actually it's an LM431

to the 900V level. It alone could control something like the FET probably.

?

As it is being used now in the "Variable HV for the LENi/CDV-700", and "Variable Bench HV Supply from a CDV-700? Transformer" in the primary side as a shunt regulator it works great, giving 200 to 1300 V stable output. I always wanted to transfer the circuit to the HV side as a shunt regulator for certain applications.

?

All sorts of current booster circuits are on the internet, nothing much about kV level, micropower bias supply use.

?

Geo

?

---

From: "Roger Whatley" <rogwhat53@...>
To: "CDV700CLUB" <[email protected]>
Sent: Tuesday, April 5, 2022 7:25:53 AM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

?

Geo,

Now, for your last question below, the answer is yes, you can replace the BJT with just the HV Mosfet and the circuit will function. In fact that is exactly Peter's original circuit, well almost...... It evolved to the subsequent forms for specific reasons though, in this thread, to improve performance.

For example, in Peter's simplified schematic the reference voltage that appears at the negative feedback terminal of the "opamp" is Vz + Vbe (crude opamp which is really a BJT gainstage with a load resistor = 1M and gain = gm*Rload). Since a 6.2V Zener has a positive tempco of about 2mV/degC and a Vbe has a negative tempco of about -2mV/degC, Wallah! (voila!).

So the BJT was a pretty big improvement in that regard, There are more comments I could make about the circuit errors with the MosFets as the gainstage instead of as the cascodes, but it suffices to say that these are what motivated the odyssey to more complicated circuitry.

However engineering economics and "good enough" motivated Peter to step back to the simpler circuit.

?

(If you had a 1000V BJT you would not need the cascodes for mere functionality.)

rogerw

?

?

Take a look at Peter's simpler schematic he just posted yesterday, and ignore the cascode MosFet, since its main function is to pass the BJT collector current up to the high voltage levels. Imagine the BJT was a 1000V Vce device and did not need the cascode for high voltage reasons.

Now, take a look at my enclosure here....... Look familiar?

rogerw

?

On 4/5/2022 4:21 AM, Geo Dowell wrote:

Really the question is, skip being variable right now for a moment, what function does the transistor play and ultimately, could you theoretically use a zener on the gate(? or where?) of the FET without the transistor to achieve a regulated output, and what value would the Zener need to be to make say, 900V?

?

Geo

?

?

?

From: "Geo Dowell" <GEOelectronics@...>
To: "CDV700CLUB" <[email protected]>
Sent: Tuesday, April 5, 2022 1:04:13 AM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

?

Peter and Roger,?

P. your hand drawn schematic helps me understand more about this.? Here's my question, with either or all of these circuits, what happens if the divider string is fixed, even if it has the added adjustments available, then the ZENER voltage is changed? Would that give the same effect??

If either one could model that, show change of HV with a range of Zener V from say 3 to 30V. And while doing that change the load resistor on P's schematic to 1.8M.

George?

---

From: "peter via groups.io" <epkoncept@...>
To: "CDV700CLUB" <[email protected]>
Sent: Monday, April 4, 2022 10:33:55 PM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

[Edited Message Follows]
[Reason: typo's]

Roger:
?Back to the circuit you posted in message #300
Simple cascode, with only 1 npn tr + 2 mosfets.
Since I found the 2SK3265, I removed the 2nd mosfet so its a very basic circuit.
66Meg, two 1 Meg, 100k+ 2.2M trimmer , 5.1V zener, 2SK3265, 2SC3311
Set trimmer for 900V, Vary input from 950~ 1150V, output only varied 5V or less.
I heated the circuit with a hair dryer, got it up from 25C to about 60C, voltage rose may be 5 or 6 volts. Good enough for me!
Its time to make it small enough to stuff inside my CDV-700 . I think the quiescent current draw while shunting was about 35uA when set to 900V with? about a 950V input.
It on the LHS of the attached jpg, circuit in message #300 is on the RHS
P

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

Virus-free.

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

开云体育

Vz is a voltage reference input to a feedback amplifer connected as a voltage gain set by feedback resistor ratio. It is in fact a positive gain opamp configuration as I showed. You can change the output voltage either by changing the Vz input OR by changing the resistor feeback ratio.

The ARE some subtle complications which lead to the conclusion that if you want it to be variable you should do as Peter shows and vary what he marked as the 2.2M resistor at the bottom of the string.

rogerw

yes that helps a lot. What is the function of V(sub)Z in determining the output V?

If the divider variable resistor is locked at a certain value, what effect does changing the Zener have on the output?

Geo

---

From: "Roger Whatley" <rogwhat53@...>
To: "CDV700CLUB" <[email protected]>
Sent: Tuesday, April 5, 2022 6:49:49 AM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

Geo,

Take a look at Peter's simpler schematic he just posted yesterday, and ignore the cascode MosFet, since its main function is to pass the BJT collector current up to the high voltage levels. Imagine the BJT was a 1000V Vce device and did not need the cascode for high voltage reasons.

Now, take a look at my enclosure here....... Look familiar?

rogerw

On 4/5/2022 4:21 AM, Geo Dowell wrote:

Really the question is, skip being variable right now for a moment, what function does the transistor play and ultimately, could you theoretically use a zener on the gate(? or where?) of the FET without the transistor to achieve a regulated output, and what value would the Zener need to be to make say, 900V?

Geo

From: "Geo Dowell" <GEOelectronics@...>
To: "CDV700CLUB" <[email protected]>
Sent: Tuesday, April 5, 2022 1:04:13 AM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

Peter and Roger,?

P. your hand drawn schematic helps me understand more about this.? Here's my question, with either or all of these circuits, what happens if the divider string is fixed, even if it has the added adjustments available, then the ZENER voltage is changed? Would that give the same effect??

If either one could model that, show change of HV with a range of Zener V from say 3 to 30V. And while doing that change the load resistor on P's schematic to 1.8M.

George?

---

From: "peter via groups.io" <epkoncept@...>
To: "CDV700CLUB" <[email protected]>
Sent: Monday, April 4, 2022 10:33:55 PM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

[Edited Message Follows]
[Reason: typo's]

Roger:
?Back to the circuit you posted in message #300
Simple cascode, with only 1 npn tr + 2 mosfets.
Since I found the 2SK3265, I removed the 2nd mosfet so its a very basic circuit.
66Meg, two 1 Meg, 100k+ 2.2M trimmer , 5.1V zener, 2SK3265, 2SC3311
Set trimmer for 900V, Vary input from 950~ 1150V, output only varied 5V or less.
I heated the circuit with a hair dryer, got it up from 25C to about 60C, voltage rose may be 5 or 6 volts. Good enough for me!
Its time to make it small enough to stuff inside my CDV-700 . I think the quiescent current draw while shunting was about 35uA when set to 900V with? about a 950V input.
It on the LHS of the attached jpg, circuit in message #300 is on the RHS
P

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

Virus-free.

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

开云体育

Geo,

Now, for your last question below, the answer is yes, you can replace the BJT with just the HV Mosfet and the circuit will function. In fact that is exactly Peter's original circuit, well almost...... It evolved to the subsequent forms for specific reasons though, in this thread, to improve performance.

For example, in Peter's simplified schematic the reference voltage that appears at the negative feedback terminal of the "opamp" is Vz + Vbe (crude opamp which is really a BJT gainstage with a load resistor = 1M and gain = gm*Rload). Since a 6.2V Zener has a positive tempco of about 2mV/degC and a Vbe has a negative tempco of about -2mV/degC, Wallah! (voila!).

So the BJT was a pretty big improvement in that regard, There are more comments I could make about the circuit errors with the MosFets as the gainstage instead of as the cascodes, but it suffices to say that these are what motivated the odyssey to more complicated circuitry.

However engineering economics and "good enough" motivated Peter to step back to the simpler circuit.

(If you had a 1000V BJT you would not need the cascodes for mere functionality.)

rogerw

Take a look at Peter's simpler schematic he just posted yesterday, and ignore the cascode MosFet, since its main function is to pass the BJT collector current up to the high voltage levels. Imagine the BJT was a 1000V Vce device and did not need the cascode for high voltage reasons.

Now, take a look at my enclosure here....... Look familiar?

rogerw

On 4/5/2022 4:21 AM, Geo Dowell wrote:

Really the question is, skip being variable right now for a moment, what function does the transistor play and ultimately, could you theoretically use a zener on the gate(? or where?) of the FET without the transistor to achieve a regulated output, and what value would the Zener need to be to make say, 900V?

Geo

From: "Geo Dowell" <GEOelectronics@...>
To: "CDV700CLUB" <[email protected]>
Sent: Tuesday, April 5, 2022 1:04:13 AM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

Peter and Roger,?

P. your hand drawn schematic helps me understand more about this.? Here's my question, with either or all of these circuits, what happens if the divider string is fixed, even if it has the added adjustments available, then the ZENER voltage is changed? Would that give the same effect??

If either one could model that, show change of HV with a range of Zener V from say 3 to 30V. And while doing that change the load resistor on P's schematic to 1.8M.

George?

---

From: "peter via groups.io" <epkoncept@...>
To: "CDV700CLUB" <[email protected]>
Sent: Monday, April 4, 2022 10:33:55 PM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

[Edited Message Follows]
[Reason: typo's]

Roger:
?Back to the circuit you posted in message #300
Simple cascode, with only 1 npn tr + 2 mosfets.
Since I found the 2SK3265, I removed the 2nd mosfet so its a very basic circuit.
66Meg, two 1 Meg, 100k+ 2.2M trimmer , 5.1V zener, 2SK3265, 2SC3311
Set trimmer for 900V, Vary input from 950~ 1150V, output only varied 5V or less.
I heated the circuit with a hair dryer, got it up from 25C to about 60C, voltage rose may be 5 or 6 volts. Good enough for me!
Its time to make it small enough to stuff inside my CDV-700 . I think the quiescent current draw while shunting was about 35uA when set to 900V with? about a 950V input.
It on the LHS of the attached jpg, circuit in message #300 is on the RHS
P

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

Virus-free.

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

开云体育

Where do I find the message number of a message? That does not apparently deliver to my email client s/w?

rogerw

Roger:
?back to the circuit you posted in message #300
Simple cascode, with only 1 npn tr + 2 mosfets.
Since I found the 2SK3265, I removed the 2nd mosfet so its very basic circuit.
66Meg, two 1 Meg, 100k+ 2.2M trimmer , 5.1V zener, 2SK3265, 2SC3311
Set trimmer for 900V, Vary input from 950~ 1150, output only varied 5V or less.
Heated the circuit with a hair dryer, got it up from 25C to about 60C, voltage rose may be 5 or 6 volts. Good enough for me!
Time to make it small enough to stuff inside my CDV-700 !
It on the LHS of the attached jpg, circuit in message #300 is on the RHS
P

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

开云体育

Geo,

Take a look at Peter's simpler schematic he just posted yesterday, and ignore the cascode MosFet, since its main function is to pass the BJT collector current up to the high voltage levels. Imagine the BJT was a 1000V Vce device and did not need the cascode for high voltage reasons.

Now, take a look at my enclosure here....... Look familiar?

rogerw

Really the question is, skip being variable right now for a moment, what function does the transistor play and ultimately, could you theoretically use a zener on the gate(? or where?) of the FET without the transistor to achieve a regulated output, and what value would the Zener need to be to make say, 900V?

Geo

From: "Geo Dowell" <GEOelectronics@...>
To: "CDV700CLUB" <[email protected]>
Sent: Tuesday, April 5, 2022 1:04:13 AM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

Peter and Roger,?

P. your hand drawn schematic helps me understand more about this.? Here's my question, with either or all of these circuits, what happens if the divider string is fixed, even if it has the added adjustments available, then the ZENER voltage is changed? Would that give the same effect??

If either one could model that, show change of HV with a range of Zener V from say 3 to 30V. And while doing that change the load resistor on P's schematic to 1.8M.

George?

---

From: "peter via groups.io" <epkoncept@...>
To: "CDV700CLUB" <[email protected]>
Sent: Monday, April 4, 2022 10:33:55 PM
Subject: Re: [CDV700CLUB] Question and an interesting site for those that modify CDV700s and more.

[Edited Message Follows]
[Reason: typo's]

Roger:
?Back to the circuit you posted in message #300
Simple cascode, with only 1 npn tr + 2 mosfets.
Since I found the 2SK3265, I removed the 2nd mosfet so its a very basic circuit.
66Meg, two 1 Meg, 100k+ 2.2M trimmer , 5.1V zener, 2SK3265, 2SC3311
Set trimmer for 900V, Vary input from 950~ 1150V, output only varied 5V or less.
I heated the circuit with a hair dryer, got it up from 25C to about 60C, voltage rose may be 5 or 6 volts. Good enough for me!
Its time to make it small enough to stuff inside my CDV-700 . I think the quiescent current draw while shunting was about 35uA when set to 900V with? about a 950V input.
It on the LHS of the attached jpg, circuit in message #300 is on the RHS
P

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

开云体育

very kewl!

Sorry I have been unresponsive this evening, been visiting with an old hunting buddy.

rogerw

Roger:
?back to the circuit you posted in message #300
Simple cascode, with only 1 npn tr + 2 mosfets.
Since I found the 2SK3265, I removed the 2nd mosfet so its very basic circuit.
66Meg, two 1 Meg, 100k+ 2.2M trimmer , 5.1V zener, 2SK3265, 2SC3311
Set trimmer for 900V, Vary input from 950~ 1150, output only varied 5V or less.
Heated the circuit with a hair dryer, got it up from 25C to about 60C, voltage rose may be 5 or 6 volts. Good enough for me!
Time to make it small enough to stuff inside my CDV-700 !
It on the LHS of the attached jpg, circuit in message #300 is on the RHS
P

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

Roger:
?I don't know the details of the avalanche process, but at low low currents ie uA's, is the temperature coefficient predictable ?

Also is the avalanche voltage knee stable ?
in other words, so the spice models include data for low current behavior ?

I vaguely remember trying to set current in the several mA range when using zeners either zener and avalanche modes.
P

?

开云体育

I have always run at least 10uA or so but in my latest circuit sim, I am running only about 4uA. Seems to work in sim-world.

Of course they are spec'd with 10mA or more........ So naturally they are running on their knees with uA.

rogerw

Roger et al:

Please refresh my mind. what sort of current was flowing through the zener(s) ??

P

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.

开云体育

Geo,

I would look forward to the same measurement but taken on a 6993 operating at 900V.

Maybe operates different from the Russkie's tube, or maybe not.....

I totally agree,wrt similarity of your result ~50V and Neon bulb..... I assume that Russian tube is a neon tube..... but 450V-50V = 400V is a long ways to slew in a (hopefully) short time. Depending on total parasitic capacitance (cabling, etc) the time it might take to do that would limit pulse rate quite a bit; I would think it would saturate at a relatively low count rate.... And maybe the 6993 is designed to improve on that???? I don't know.... I may look for some more datasheets for comparison.

I definitely don't understand everything I know at this point.

rogerw

Ah good. Somehow this important Topic got?"locked" last night. If a moderator did that please let me know. It's unlocked again.

Here's the latest and more critical look at the Bias HV drop across the SBM-20 GM tube operated at 450V via a re-regulator base from 920V DC from a Ludlum Bench "Geiger Counter".

First the last night report- updated?

"While attempting to clean up the workbench today, I had to look into this. Using a Russian SBM-20 (because it was on the workbench for a different project) and one of my 900 to 450V adaptor bases, scope attached directly across the GM tube (scope rated for 2000V), and a mantle for a source, the peaks on the scope showed negative pulses measuring f 450V P-P. At the "Geiger Counter" actually a (Ludlum 2500 Bench scaler) the P-P was 900V."

Looking again this morning with caffeinated eyeballs, a selection of eyeglass and more patience. Specifically looking to see if the "valley" goes all the way to zero before resetting. Best I can tell it does not go to zero, but the exact floor is difficult to determine with this 'scope, it seems to be somewhere around or just above 50V above ground. This would track very well with the NE-2 Neon Bulb analogy.? Someday (not this day for sure), we'll nail that down, with the storage 'scope, or someone else can.

By the way a GM tube activates with radiation because it is being operated on the knee of the gas discharge tube curve, in the Geiger region, and there is always a wide latitude of exact HV settings that allow that proper operation. + or - 50V or even more is normal, and is what Rad Techs call the "plateau" area (flat spot) of the curve.

Back to the discussion guys, sorry for the unintentional temporary lockout of us all.

Geo

-- 
rogerw
On the Catclaw
ΜΟΛΩΝ ΛΑΒΕ

The PanDemic is over, but the DemPanic goes on.
MisDisMal is a dictator's tool of repression.